So I took a look at the two boards' solder joints; particularly all the capacitors. I really coundn't spot anything ecan resembling a hairline crack in any of them. But I did notice something fishy.
Regarding what @JoeO said, it seems this board has a very convoluted way of getting grounded. Theres a screw on the board that connects a metal bracket to the board; and that metal bracket has thin fins that are supposed to touch the metal backplate. The metal is really thin and all bent out of shape, no doubt due to thermal cycling. What a terrible design!
I also found a lot of crud/oxide buildup between the screw and the bracket. I cleaned that off and bent the brackets in an exaggerated amount to force a good connection.
Now the screen powers up and switches channels much faster than before, and actually now just as fast as the other two! Ill see if the screen powers down in the next few weeks but now its performing better than the day I got it, so Im optimistic! Looks like fretting / oxide buildup was the likely cause, the screens weren't new so I suppose they could have been thermal cycled many times by the time I got them. Also, these screens seem to have verry underpowered PSU's and are susceptible to small ground fluctuations. When I power anything on thats on the same line as these screens, the screens all go dark for a second. So I suspect a poor contact could make them float intermittently. What a heap of junk but they were cheap!
Thanks for the help!